Reconciling model-X and doubly robust approaches to conditional independence testing

Z Niu, A Chakraborty, O Dukes… - The Annals of …, 2024 - projecteuclid.org
Reconciling model-X and doubly robust approaches to conditional independence testing Page
1 The Annals of Statistics 2024, Vol. 52, No. 3, 895–921 https://doi.org/10.1214/24-AOS2372 © …

Differentially Private Permutation Tests: Applications to Kernel Methods

I Kim, A Schrab - arXiv preprint arXiv:2310.19043, 2023 - arxiv.org
Recent years have witnessed growing concerns about the privacy of sensitive data. In
response to these concerns, differential privacy has emerged as a rigorous framework for …

Dimension-agnostic inference using cross U-statistics

I Kim, A Ramdas - Bernoulli, 2024 - projecteuclid.org
Additional results are provided in the supplementary material [43]. Appendix A discusses
multiple sample-splitting, while Appendix B describes a general strategy for studying the …

The Minimax Rate of HSIC Estimation for Translation-Invariant Kernels

F Kalinke, Z Szabo - arXiv preprint arXiv:2403.07735, 2024 - arxiv.org
Kernel techniques are among the most influential approaches in data science and statistics.
Under mild conditions, the reproducing kernel Hilbert space associated to a kernel is …

Dimension-agnostic inference using cross U-statistics

I Kim, A Ramdas - arXiv preprint arXiv:2011.05068, 2020 - arxiv.org
Classical asymptotic theory for statistical inference usually involves calibrating a statistic by
fixing the dimension $ d $ while letting the sample size $ n $ increase to infinity. Recently …

Robust Kernel Hypothesis Testing under Data Corruption

A Schrab, I Kim - arXiv preprint arXiv:2405.19912, 2024 - arxiv.org
We propose two general methods for constructing robust permutation tests under data
corruption. The proposed tests effectively control the non-asymptotic type I error under data …

Half-KFN: An Enhanced Detection Method for Subtle Covariate Drift

B Wang, D Xu, Y Tang - arXiv preprint arXiv:2410.08782, 2024 - arxiv.org
Detecting covariate drift is a common task of significant practical value in supervised
learning. Once covariate drift occurs, the models may no longer be applicable, hence …

Studentized Tests of Independence: Random-Lifter approach

Z Gao, R Wang, X Wang, H Zhang - arXiv preprint arXiv:2410.18437, 2024 - arxiv.org
The exploration of associations between random objects with complex geometric structures
has catalyzed the development of various novel statistical tests encompassing distance …

Efficiently Learning Significant Fourier Feature Pairs for Statistical Independence Testing

Y Ren, Y Xia, H Zhang, J Guan, S Zhou - The Thirty-eighth Annual … - openreview.net
We propose a novel method to efficiently learn significant Fourier feature pairs for
maximizing the power of Hilbert-Schmidt Independence Criterion~(HSIC) based …